categorise all the ZOWE components under sensible report classes, and look
at the 72 GOAL SMF to get an aggregated view?
On Sat, 11 Jan 2025 at 20:08, Colin Paice <
059d4daca697-dmarc-requ...@listserv.ua.edu> wrote:
> I'm running ZOWE. The started task kicks off various SH commands. In
Keep subscriptions to an absolute minimum.
ideally just one.
On Wed, 10 Jul 2024 at 12:14, Tommy Tsui <
069a23f5affb-dmarc-requ...@listserv.ua.edu> wrote:
> Hi all
>
> Any shop use CDC for db2 capture high volume and high frequency update Db2
> table to kafka? Any pros and cons?
>
> Thanks
>
"It doesn't take an extremely large number of CPUs
before a single-image system will deliver less capacity than a sysplex
configuration of two systems, each with half as many CPUs".
In the original context of the GSE material, does "system" here mean
physical CEC, or LPAR?
It is unclear, and "It
A few years back, I did a deep dive into tuning CPU usage across a
multitude of very small z/OS guests under z/VM, and WLM was certainly a big
hitter for many of them, but as there were so many instances, I was able to
see notable differences in WLM use between "LPARs", which was obviously "of
inte
I have seen Db2 unloads use vastly more (overall total) CPU when doing
parallelised processing, compared to non-parallelised.
Obviously you are doing reload (I presume) rather than unload, but the same
thing may apply.
So might be something worth checking/testing, if parallelism is relevant to
your
try this ODS example
http://support.sas.com/kb/23/652.html
On Wed, 2 Aug 2023 at 22:25, Horne, Jim wrote:
> I just googled to find it. I haven't done a .CSV myself in ages but it
> can be done. You could use PROC SQL to create a macro variable then a
> dummy data step to write it to a dataset
ISMF option 2 can get you a list of all your volumes.
Whilst not %ages, FREE DSCBs and FREE VIRS will give you an idea of what
you might be running low on.
Columns are sortable.
Use DCOLLECT if you want to do things in batch.
On Wed, 24 May 2023 at 15:39, Steely.Mark wrote:
> Received this m
https://www.lookupmainframesoftware.com/soft_detail/dispsoft/162
On Tue, 25 Apr 2023 at 13:01, rpinion865 <
042a019916dd-dmarc-requ...@listserv.ua.edu> wrote:
> Take a look at
>
> https://cvsystems.com/cv-payment-processing/
>
>
>
>
> Sent with Proton Mail secure email.
>
> --- Original
SMF 119 ZERT records have a significant amount of potentially relevant
information, assuming they are turned on, and you are able to format them.
On Wed, 14 Dec 2022 at 19:58, Crusty Old Guy
wrote:
> Yes, AT-TLS is involved. This is an HTTPS call from Chrome to a port in a
> CICS region. The r
For each I/O metric, if you divide the value by the average, and multiply
by 128, then you roughly get the IOrate.
It is the same on our systems. always 128.
Whether it is intentional, or there are some wrong sums being done
somewhere, i cant say.
On Sat, 26 Nov 2022 at 09:43, Boesel Guillaume
is the delay at regular intervals, or random?
If regular intervals, then that may give some clues.
On Sat, 27 Aug 2022 at 18:30, Peter wrote:
> Steve
>
> It's a non CICS application
>
> On Sat, Aug 27, 2022, 9:18 PM Steve Beaver wrote:
>
> > You are going to hate this but you need to look at yo
SMF 92 subtype 11 is usually useful in showing USS file activity, which can
tend to reveal if anything 'odd' is going on with file based activity.
I/O is also seen when network activity is involved, but I have not found
92's very useful in determining what is going on in that arena, but others
may
30 years ago, in the early days of SMS, when all the "logical space stuff"
appeared, It crossed my mind that it might be "nice" to have a construct
whereby space could be based on specifying the number of records, rather
than a particular logical or physical space quantity.
e.g. SPACE=(RECS,(20,10)
Another option:
https://ftpdocs.broadcom.com/cadocs/d0/d93e.pdf
On Fri, 30 Oct 2020 at 22:51, Tony Thigpen wrote:
> All,
>
> I have a z/VSE client that believes it is time to move to z/OS. But,
> they have one big concern. They have a lot of ShareOption=4 VSAM files.
>
> For those that do
Also there is the ISPF built in ISRDDN.
ENQ command on the command line.
On Thu, 15 Oct 2020 at 05:32, Brian Westerman
wrote:
> Also there is MXI (option ENQ). Also on the CBT
>
> --
> For IBM-MAIN subscribe / signoff / archive
...and if it's an SMS volume, you also need to concern yourself about VVDS
size.
On Sun, 5 Jul 2020 at 12:56, Mike Schwab wrote:
> Here are the VTOC sizes needed for volumes full of 1 track datasets.
>
> https://www.ibm.com/support/knowledgecenter/en/SSLTBW_2.1.0/com.ibm.zos.v2r1.ickug00/ick407
If you are only interested in which datasets are on which volumes, then
DCOLLECT would surely do the job.
On Mon, 22 Jun 2020 at 15:45, Brian France wrote:
> Many thanks to all the responses!!!
>
> David,
>
> Thank you!!! This is exactly what we are looking for. I just
> changed a sort to
ASM2 used to tuck things away in the VTOC via an exit. Last modified
jobname & last modified date are the two fields i specifically remember
being present in our setup. $RSVP could then be used to report on these
"unofficial" fields.
On Mon, 1 Jun 2020 at 13:48, Allan Staller wrote:
> There i
We found higher CPU was generally down to the new version of LE routines.
COBOL itself was not found to be an issue CPU-wise, although there was only
limited comparative testing done.
On Thu, 21 May 2020 at 13:48, Pommier, Rex wrote:
> OPT(0) burns that much more CPU? Is this on all compiles or
Some scrutiny of the DFHSM activity logs to see if the same old datasets
are failing to be processed each time, may not go amiss. That sort of
stuff can build up over time, and can accumulate to become a considerable
overhead every time automatic functions run.
On Tue, 28 Apr 2020 at 06:50, Bria
Your Space constraint relief definitions in your data classes will affect
things substantially, depending on what your SMS routines are assigning,
and what settings your dataclass has.
https://www.ibm.com/support/knowledgecenter/en/SSLTBW_2.3.0/com.ibm.zos.v2r3.idad400/scra.htm
On Thu, 21 Nov 2
We have unique (off-host) application servers back-ending a mobile app,
which sends requests into our CICS ecosystem using unique CICS userids,
which is how we separate them out (as there is massive reuse of exactly the
same transactions between numerous applications in our environment) for
MWP, us
I have never liked the distraction of a crosshair cursor.
The one exception was 3 decades ago, using a 4-in-1 3290 screen in the ops
bridge, where it was sometimes not easy to keep track of which of the 4
screens was 'active', which the crosshair did actually make it a lot more
obvious than the us
You may want to look at CA Shareoption5. I know that was used in a
VSE->z/OS migration in a past life, and seem to recall it was something to
do with addressing online & batch concurrency/integrity.
On Wed, 5 Sep 2018 at 23:24, Mark Waterbury <
01c3f560aac1-dmarc-requ...@listserv.ua.edu> wro
When AVGREC/AVGVAL came about in the early SMS era, I thought, given our
generally records-based view on things in the z/OS world, it would be
'nice' if logical space allocations could be done effectively requesting
numbers of records (as an alternative to having to explicitly request
bytes).
Thus
Owner is SDSFAUX in our env.
On 19 April 2018 at 19:58, Martin Packer wrote:
>
> Looking at the member and job name fields in SMF 74-2 might well give you a
> clue.
>
> It’s not one I’ve seen bubble up when I analyse 74-2 (which I do
> practically any engagement).
>
> Cheers, Martin
>
> Sent fro
I have 'rolled my' own job flow diagrams in the past, augmenting with SMF
data to allow CPU (or elapsed) time to determine the size of the plotted
boxes to be able to visualise the heavy hitters in the flow.
I had two iterations:
The first generating SGML that could be subsequently input to yED to
John, are you specifically after a 'graph database', or a 'pure java'
database?
H2 is an example of a pure java database, which is >99% zIIP eligible, in
my experience.
Any pure java library (typically in the form of jar file[s]) should 'just
work' on z/OS (i.e. no porting required).
Havent quit
5-dmarc-requ...@listserv.ua.edu> wrote:
> > I verified the MC and DSORGs. No problem found. No patches either
> which would impede the PSM. I remember reading or being told that once the
> SM conditions had been met, HSM stops looking at the volume(s).
> >
> >
Primary space management should normally process all volumes in a storage
group. And PSM should normally look to reduce utilisation of a volume down
to the low threshold (caveat - various PATCHes can alter DFHSM standard
behaviour).
Have you checked the messages for space management actions for the
Whilst you're waiting for IBM to deliver a proper API, you could trap the
console output from a F DFHSM,Q CDS.
DFHSM may well not be called DFHSM of course, so that will add a little
extra work, perhaps using an ISGQUERY on ARCENQG
On 17 January 2018 at 13:29, Manfred Lotz wrote:
> On Wed, 17
ain aim at least initially, is to get a view of the WEB queue totals at
least once a second, so this doesnt really seem to fit the bill.
On 27 November 2017 at 00:06, Roger Lowe wrote:
> On Sun, 26 Nov 2017 01:33:40 +, Graham Harris
> wrote:
>
> >Would anyone happen to
Would anyone happen to know where the "SYSLOAVG" (stands for 'system load
average') value in RMF3 comes from?
I think it is also known as the WEB queue.
I have ploughed through the data areas manuals, and cant see any obvious
candidate control block at an LPAR level, but i may not be searching for
If you have the RMF DDS server (GPMSERVE) running, the you can access this
sort of information from the RMFIII data sources via http requests.
e.g. a list of actual MSU values across all accessible partitions, would be
something like:
youripaddr:8803/gpm/perform.xml?resource="yourplex,SYSPLEX"&id
>o Is there a z/OS utility to generate a CCKD? Or, would it be necessary
to run
> Hecules in a Linux guest to which the source volume could be ATTACHed?
http://www.bsp-gmbh.com/turnkey/cookbook/hercules/cckddasd.html#cckddump
On 12 October 2017 at 17:26, Paul Gilmartin <
000433f07816-dmarc-
Websphere bods need to turn SMF on on their side.
On 27 April 2017 at 22:18, Blaicher, Christopher Y.
wrote:
> Can you give us the output from a D SMF,O command?
>
>
> Chris Blaicher
> Technical Architect
> Mainframe Development
> P: 201-930-8234 | M: 512-627-3803
> E: cblaic...@syncsort.com
>
Have you got RMF3?
Tried looking at STORF to see what address space(s) are consuming all your
AUX slots?
Are you running Netview? Got a CANZLOG dataspace? That can chew through a
lot of memory if not configured quite right in recent z/OS releases.
On 13 April 2017 at 20:26, Jesse 1 Robinson w
ah - after looking at your JCL again, a no DSN= specification won't work in
your example, as you do need the &&SYSEXEC in order to specify the member.
On 11 March 2017 at 01:23, Graham Harris wrote:
> >>100 records written & read inbetween alloc & free:
> >&g
a temporary dataset if you do not specify a DSN=
parameter at all. Which should work in your particular example i think.
On 11 March 2017 at 00:18, Paul Gilmartin <
000433f07816-dmarc-requ...@listserv.ua.edu> wrote:
> On Fri, 10 Mar 2017 23:46:07 +, Graham Harris wrote:
>
&
100 records written & read inbetween alloc & free:
DASD:5.5 mins elapsed 92 secs CPU
VIO 73 secs elapsed 65 secs CPU
On 10 March 2017 at 22:48, Tom Marchant <
000a2a8c2020-dmarc-requ...@listserv.ua.edu> wrote:
> On Fri, 10 Mar 2017 21:05:20 +0000, Gr
Being somewhat interested in validating whether VIO still serves a purpose
in these modern times with super quick DASD, i just ran some mickey mouse
comparisons, doing 10k allocates/frees of temporary datasets on VIO vs DASD.
Results tend to imply that the old adage of "the best I/O is no I/O", st
Mobile workload reuses a reasonable chunk of our existing CICS capability,
and shares the same regions as everything else. Unique userids in CICS
allows the mobile transactions be be fully identified (userid, usefully,
propagates through the whole UOW starburst). An hourly summarisation of
CICS 11
Might employing the GRS enqueue monitor help understand what is going on
behind the scenes, and in what order?
On 7 December 2016 at 00:27, Kirk Wolf wrote:
> There are two SPFEDIT enqs required to follow the protocol. The data set
> one (without member) is used to serialize during actual I/O.
You could save yourself a lot of trouble and do a SAVESCDS before you make
your ACS routine changes.
On 12 October 2016 at 18:08, Chuck Kreiter wrote:
> Thanks for the reply.
>
> The problem I'm having is that the definition for each volume as defined
> in the storage group doesn't show up o
Out of curiosity, I have just tested my earlier suggestion of using DFDSS
logical dump, to dump a zEDC compressed dataset.
The resultant dump dataset is as small as the z/EDC compressed dataset,
which is what i had anticipated.
I then transferred it to another sysplex, restored it, and it appears t
Interesting question!
My first thought would be to do a DFDSS logical dump of it.
You would probably then need to terse it if you want to send via FTP.
Would be interested in the results.
On 29 September 2016 at 14:09, Mauri Kanter wrote:
> Hi:
>
> My understanding is that one a file was com
Something to be aware of when looking at CPU delays being reported by
RMFIII, is that in multi-tasked asids, if ANY of the subtasks is found to
be in CPU wait when RMFII does it sampling sweep (once per second), then
that sampled second will have a 'CPU wait' associated with it, applying to
that as
As we have no plans to use HIS sampling (which I presume is what makes use
of the data), it seemed an unnecessary waste of real memory. It is a
decision that is easily reversed if this data may become valuable in the
future.
On 21 August 2016 at 23:45, Peter Relson wrote:
> >we also have turned
Upped ours to 200Gb, after the tenth DB2v10 subsystem on one of our Dev
LPARs refused to start up.
Whilst on the subject of HVCOMMON, we also have turned off TRACKDIRLOAD (in
PROGxx) after seeing measurable growth in HVCOMMON usage after z/OS2.1
upgrade, the connection between the two only being c
Did someone change your VTS replication policy? Sounds like it is now
having to wait for the volume to be fully transferred to another VTS to
assure you have two copies, before the rewind/unload is allowed to
complete.
https://www.ibm.com/support/knowledgecenter/SSLTBW_2.1.0/com.ibm.zos.v2r1.idao3
With our VTS, we found ourselves in the position where the default
automatic DFHSM recycle was unable to keep up with scratch cartridge
release that was needed to maintain the required scratch count, which was
putting the whole VTS environment at risk (90+% of which was consumed by
DFHSM migration)
Ed Jaffe wrote:
>The nice thing about an overabundance of real storage is the ability to
create a very large LFAREA with plenty of room for 1MB >pageable pages and
-- if you have software that can truly benefit from it (which we don't) --
2G fixed pages.
LFAREA does of course only generally hol
Tom Brennan wrote
>
> cron tasks will be a problem I think, since a non-priviledged user can
> setup their own private "autocommands" on unix. Not so on the mainframe.
> But from what I've seen, unix people tend to want cron tasks because they
> are polling for data, which needs to be discouraged
Do you have another dataclass you can use which has Space Constraint Relief
= YES?
On 4 February 2016 at 17:25, Mike Geiger wrote:
> Additional information. The SCDS does match the ACDS(I really did do a
> setsms, many of them). Had I been thinking a bit deeper I would have used
> the ACDS fo
I noticed a common DUMP queue mentioned on p14 of this presentation,
No sign of CMQ though..
https://share.confex.com/share/125/webprogram/Session17814.html
On 23 November 2015 at 15:02, Lizette Koehler
wrote:
> I have been hearing rumors that in a future release that DFHSM will have
> somethi
In z/OS V1R12, DFSMS access method services (IDCAMS) adds a new wildcard
option to the DELETE command, which lets you delete all members of a PDS or
PDSE.
https://www-01.ibm.com/support/knowledgecenter/SSLTBW_1.13.0/com.ibm.zos.r13.idak100/amspdse12.htm%23amspdse12
On 19 November 2015 at 21:42,
MXI has comprehensive ENQ query capabilities, and a very usable REXX
interface as well.
On 25 September 2015 at 01:44, Anthony Thompson
wrote:
> For trapping the results in a REXX you could use the SDSF ISFSLASH service
> to capture the results of a D GRS, RES= or D GRS,C command. I imagine you
MXI is another option.
On 31 May 2015 at 14:19, Shmuel Metz (Seymour J.) <
shmuel+ibm-m...@patriot.net> wrote:
> In <55673d3b.2020...@vse2pdf.com>, on 05/28/2015
>at 12:07 PM, Tony Thigpen said:
>
> >Is there simple batch method to get a list of all on-line dasd
> >volumes without actually
If you have MXI, then you may find SGRP gives you what you want.
On 24 March 2015 at 22:22, Shmuel Metz (Seymour J.) <
shmuel+ibm-m...@patriot.net> wrote:
> In
> ,
> on 03/24/2015
>at 04:51 PM, Mike Schwab said:
>
> >IDCAMS DCOLLECT by volume into a file. LSPACE Macro for in program
> >use.
Would checking the DDNAME and/or PROGRAM help?
On 17 March 2015 at 15:10, Vernooij, CP (ITOPT1) - KLM <
kees.verno...@klm.com> wrote:
> Well, the temp-dsn pattern will probably only change with z/OS releases,
> so it will (only) be another item on that checklist.
>
> Kees.
>
> -Original Messa
it was the whole shooting match.
http://sortbenchmark.org/
'rules' are here http://sortbenchmark.org/FAQ-2014.html
On 18 January 2015 at 15:15, Staller, Allan wrote:
> What was the elapsed time for IO.
>
> I.e. The sort ran in 23 min, how long did it take to read and write the
> file?
>
> The
Impressive indeed.
I wonder how quick a fully configured z13 could do it in.
Sadly, i suspect we will never know.
On 16 January 2015 at 18:37, John McKown
wrote:
> http://opensource.com/business/15/1/apache-spark-new-world-record
>
> In October 2014, Databricks participated in the Sort Benchmar
Crikey - we replaced 3380K's with 3390-2's back in 1990. We then had to
replace the whole lot again soon after with 3390-3's, but thats a whole
other story...
On 4 December 2014 at 23:31, Tom Brennan wrote:
> Joel Ewing wrote:
>
>> Every MVS volume I have seen in the last two decades is on an e
and they arent doing a very good job;)
http://www.reveal.co.uk/real-life-stories/news/a613837/student-receives-over-gbp3k-of-gifts-from-amazon-thanks-to-tech-glitch.html
On 2 December 2014 at 20:43, Ed Finnell <
000248cce9f3-dmarc-requ...@listserv.ua.edu> wrote:
> _These Cool Robots Are
I setup a filter in gmail to send all IBM-MAIN stuff to its own folder (or
label?), and my experience is that nothing (IBM-MAIN-wise) then gets sent
to spam - the filter overrides it.
On 30 November 2014 at 08:07, Ed Finnell <
000248cce9f3-dmarc-requ...@listserv.ua.edu> wrote:
> It's at the b
tic purposes.
I expect there is a certain amount of application 'noise' in there, which
is irrelevant for system purposes, but may be quite important for the
application folks (again, perhaps only in retrospective diagnosis of
issues).
On 24 November 2014 at 21:06, John McKown
wrote:
> O
25million
On 24 November 2014 at 18:07, Greg Shirey wrote:
> I'm surprised, but it looks like on an "average" day the SYSLOG generates
> 1.3 million lines.
>
> Regards,
> Greg Shirey
> Ben E. Keith Company
>
> -Original Message-
> From: IBM Mainframe Discussion List [mailto:IBM-MAIN@LIST
I have in the past, used CA-MIA to share tape drives across a number of
sysplex/monoplexs.
On 10 November 2014 22:20, Shane Ginnane wrote:
> On Mon, 10 Nov 2014 16:01:57 -0600, John McKown wrote:
>
> >But I would still go with a CTC based _Basic_ Sysplex.
>
> Not these days - XCF based. Works
I think you need to supply a bit more information about what exactly you
are doing, and/or expecting to happen. I notice you have a VERIFY(FC1800)
on your INIT. So, the FC1800 volser was originally NONSMS? You then
re-initialise as SMS, with the same volser? You then add it into a Storage
Grou
Did you look at DCOLLECT?
That has all the fields you want.
This is a z/OS2.1 link to the fields, but i am sure you have access to your
own appropriate docs.
http://www-01.ibm.com/support/knowledgecenter/SSLTBW_2.1.0/com.ibm.zos.v2r1.idai200/dgt3i294.htm
On 30 July 2014 16:50, Christian Birr
wr
The way we load things into cache is to use the BVIR logical-to-physical
mapping report to build JCL streams to reference the logical tapes to bring
them into the cache, with a single physical mount. Can't help you with
removing from the cache, but would interested if you do find out! Thought
i d
I recall Storagetek talking about RAIT around a decade ago. Not sure if
anything came of it in the mainframe space. The concept seems to be 'out
there' on other platforms.
On 1 June 2014 01:41, Mike Schwab wrote:
> There is Error Correcting Code to help recover each block. Up from
> Parity
Use of RMF Data server (GPMSERVE) may be an option.
On 9 May 2014 23:37, Roger Lowe wrote:
> On Fri, 9 May 2014 18:13:05 +, ITURIEL DO NASCIMENTO NETO <
> 4254.itur...@bradesco.com.br> wrote:
>
> >I'm doing an assembler program that needs to know actual CPU consumption
> of every Lpar in th
You might want to check your VTOC Index's havent got trashed on some of
your volumes. Without an enabled VTOCIX, SMS will never update its view of
free space info for that volume (I think it just continually sees what it
had free the last point in time the VTOCIX was enabled).Cant work out
if
Might be worth trying ISMF. I have found it amazingly adept at deleting
'broken stuff' in the past.
On 4 February 2014 20:01, Mike Schwab wrote:
> We have an orphaned VVR in a VVDS. Not in the catalog or VTOC.
> DELETE data.set.name FILE(DD1) VVR failed NOT CATALOGED.
> What do we need to do
What is your SDSP threshold value? (i.e. the SDSP= value on a query setsys)
On 29 January 2014 21:30, Mike Schwab wrote:
> Yes. They were all INDEX(0,1,074) VTOC(05,00,1500).
> When we get them empty we will re-init as INDEX(0,1,149) VTOC(10,00,1305).
> Yesterday this LPAR had 192,000 HSM log
http://sourceforge.net/projects/universalcobol/
On 22 January 2014 22:41, Graham Hobbs wrote:
> What of GNU COBOL? Is free.
> Graham Hobbs
>
>
> On 22/01/2014 4:48 PM, Frank Swarbrick wrote:
>
>> I imagine one could use Micro Focus COBOL and its JVM support. I suppose
>> mainframe file I/O mig
Could additional potential CPU efficiencies be gained from creating the
memory objects with 1Mb large pages (or 2Gb if they are REALLY big!)?
On 20 January 2014 22:40, John Gilmore wrote:
> Jim.
>
> You also meant TARGET_VIEW=HIDDEN, not TAGET_V IEW=HIDDEN.
>
>
> John Gilmore, Ashland, MA 01721
Some (most?) sites have a policy of "putting beyond use" storage
devices/media that need to leave their data centre confines for some reason
(e.g. failure-replacement, end-of-life).
Simplistically, this could be via physical destruction (crushing, platter
hole punching, sledgehammer, etc.) or logic
or if you really do have CPU cycles going spare, at least use TIME('E') !
On 14 January 2014 21:42, Paul Gilmartin wrote:
> On Tue, 14 Jan 2014 16:12:31 -0500, Jim Mulder wrote:
>
> > There have been a few discussions over the years on IBM-MAIN as to
> >how to SLEEP for some period of time in
There are no special SMF records for java stats per se. I must say, that
the amount of CPU indicated in your table, is somewhat surprising, as java
simply isnt that 'cheap'. Especially, when considering the elapsed
timejava very rarely sits around NOT consuming CPU when it has 'things
to do'!
here's one consideration, which may or may not be an issue for you.
http://www-01.ibm.com/common/ssi/cgi-bin/ssialias?htmlfid=897/ENUS913-145&infotype=AN&subtype=CA&appname=skmwww
On 7 October 2013 21:44, Ward, Mike S wrote:
> Hello all, can any of you tell me why we would not want a z114 ove
September 2013 13:56, Robert A. Rosenberg wrote:
> At 21:40 +0100 on 09/18/2013, Graham Harris wrote about Re: HMIGRATE in
> parallel:
>
>
> I agree. What should not be hard for HSM to create multiple migration
>> subtasks (each with its own ML2 dataset) and have the master tas
I agree. What should not be hard for HSM to create multiple migration
subtasks (each with its own ML2 dataset) and have the master task pass one
dataset at a time to each subtask (ie: The subtask does the same work as is
currently done for a migrate). As each subtask finishes it asks the master
ta
Perhaps some clarification of where the 'large memory numbers' are being
reported from by the OP.
In my site, looking at *MASTER* REAL usage, both in SDSF DA and RMF3 STORF,
appear to account for real memory frames held by dataspaces owned by that
ASID (which makes sense, as where else would they b
Martin, how about Dataspaces? *MASTER* does own quite a few of them!! One
of our systems currently has 1.2Gb lodged against MASTER, which whilst not
quite 2Gb, is 'up there' compared to most other LPARs in the plex.
On that LPAR, CSM64001 accounts for 190k frames of REAL, which is best part
of 80
Although is a little hassle to setup if you dont have it already, there is
the SMS storage class exit, which can be found in SYS1.SACBCNTL(IGDACSSC).
I used this a lot in the early days of SMS, specifying ACCT=DEBUG on the
EXEC card of the job/STC/TSO-id of the task in question, to prompt the exit
In the early 90s, I came close to raising a share requirement to allow the
AVGVAL specification to default to picking up the lrecl (i.e. when not
specified explicitly) such that it would effectively allow the space request to
specify a quantity of records rather than a byte count. That would ha
I understand the root cause of this issue (which is essentially repetitive
parallel execution of large numbers of SRBs within ANTMAIN when
flashcopy eligibility/pairing checking is applied against very large
numbers of volumes) is "recognised", and is intended to be addressed in a
future version of
6 1/2 years here
On 11 May 2013 15:04, Ed Jaffe wrote:
> I just noticed this insert added to msgIXC334I by OA41210:
>
> SYSPLEX INITIALIZATION TIME: 12/29/2011 20:59:16.385348
>
> So this sysplex is about 1 1/2 years old. Are there production sysplexes
> that are decades old?
>
> --
> Edward E
We too have had issues in the past with "high CPU" in ANTMAIN, although
generated by DFDSS COPY rather than BMC utilities, but the root issue could
be similar, or the same.
The problem for us was eventually narrowed down to specifically occurring
when allocations were targetting (in our case) SMS s
As mentioned already, "its all about the TLB". With memory intensive tasks
using large pages, it lessens the usage of TLB from what would otherwise be
the case, and that potentially benefits (by DAT avoidance) not only the
task using large pages, but also every other workload utilising that TLB
(w
ISMF is (unusually) helpful in getting rid of such orphan/duff datasets,
avoiding scrabbling around trying to sort out the correct JCL.
On 26 March 2013 12:37, Tom Marchant wrote:
> On Mon, 25 Mar 2013 17:21:00 -0400, Robert A. Rosenberg
> wrote:
>
> >At 14:00 -0500 on 03/25/2013, Paul Gilmartin
If you could tolerate the pain/overhead of creating an ISPF environment for
your batch rexx to run within (or are you already doing this anyway, if you
expect to VPUT items?), then you could use the builtin QUERYENQ ISPF
service.
http://publib.boulder.ibm.com/infocenter/zos/v1r12/index.jsp?topic=%
Sorry, correct link:
http://publib.boulder.ibm.com/infocenter/zos/v1r12/index.jsp?topic=%2Fcom.ibm.zos.r12.idas200%2Fs2010.htm
On 9 January 2013 20:03, O'Brien, David W. (NIH/CIT) [C] <
obrie...@mail.nih.gov> wrote:
> Don,
>
> Please refer to 11.6.5 ACS routines invoked for restoring, recalling
Dataclass ACS is ONLY invoked on initial dataset allocations
(&ACSENVIR='ALLOC') - possibly with some very exceptional exceptions (I'm
not bang up to date on possible recent developments in SMS-land).
But this link confirms categorically that HSM recalls and numerous other
'ACS environments' do no
The technique I used from the outset of SMS (early 90's), was to deploy
QUINEW volumes either in the storage groups of interest, or a common/shared
OVERFLOW storage group that your SG ACS routine would point to for the
storage groups of interest.
This does rely on appropriate HI-ALLOC thresholds b
Google is your friendkeyword: HiperDispatch
On 10 December 2012 23:13, גדי בן אבי wrote:
> Hi,
>
> We just upgraded to z/OS 1.13
> When we look at the processor status, we see that two of the four
> processor are marked parked.
> We see this using CA-SYSVIEW v13.5 – the CPU command.
>
> W
On 30/10/2012, Ron Thomas wrote:
> LRECL is fixed. I believe kolusu solution will work.
>
> Thanks,
> Ron T
>
> --
> For IBM-MAIN subscribe / signoff / archive access instructions,
> send email to lists...@listserv.ua.edu with th
do a google on 'yed'
On 28/10/2012, Roberto Ibarra Magdaleno wrote:
> Sorry that should say HCM (Hardware Configuration Manager) not HMC…
>
>
> On Sun, Oct 28, 2012 at 2:57 PM, Roberto Ibarra
> wrote:
>
>> Hello everybody,
>>
>> When you print an HCD Graphical Configuration Report to a GML file,
1 - 100 of 102 matches
Mail list logo
